Introduction

Obesity is a problem of epidemic proportions [1]. In the USA, nearly two thirds of the population is overweight or obese, and obesity-related conditions such as cardiovascular disease, obesity-related cancers, and diabetes are among the leading causes of preventable death in the nation [2]. Although obesity incidence rates are slowing, prevalence remains high, and current trends suggest that three fourths of the population may be overweight or obese by 2020 [3].

Obesity is a multifactorial disease with influences from individual, social, environmental, and “macro-level” factors [4••, 5•] (Table 1). Individual contributors to obesity include biology (e.g., genetic, endocrine, and metabolic disorders), behavior (e.g., excess calorie consumption and limited exercise), and psychology (e.g., depression, stress, and anxiety). Social contributors include the role of culture, family, friends, and peers in creating social norms and role modeling behaviors, as well as socioeconomic factors (e.g., race, income, and education). Environmental contributors include access to healthy foods, energy-dense (e.g., fast food) foods, and safe walking spaces in one’s neighborhood. “Macro-level” factors include the influence of media and advertising, access to healthcare, and various social, economic, and government policies. Together, these factors have shaped the obesity epidemic and serve as potential targets for obesity management.

Table 1 Individual and non-individual domains of obesity: example causes and management strategies
Full size table

Research has shown the importance of managing both individual and non-individual contributors to obesity [6]. For example, weight loss interventions that include social support from family and friends have shown greater weight loss among those who lose weight with a partner (vs. those without such social support) [7]. Additionally, food store interventions that promote healthy food options have shown positive effects on individual weight; and culturally tailored, community-based interventions that promote a non-obesogenic environment have resulted in significant weight loss on a population scale [8]. Multi-level approaches to obesity management have produced improvements in both proximal (e.g., exercise and food-related behaviors) and distal obesity outcomes (e.g., body mass index) and have formed the basis for community-wide recommendations to increase access to affordable healthy foods, decrease access to unhealthy foods, and promote access to safe spaces for physical activity [9]. Perhaps the most well-known multi-level obesity intervention was the North Karelia Project that utilized community-based partnerships, marketing, and governmental policy to improve diet and reduce cardiovascular mortality among Finnish citizens from the 1970s through the 1990s [10].

Unfortunately, research on how to translate these well-known public health insights into clinical practice has been lacking. While there is some research that offers suggestions for how physicians can address the non-individual causes of obesity (e.g., screen for socioeconomic issues, coordinate services for individual patients by partnering with social workers, health advocates, community health workers, and similar professionals), this research currently represents expert opinion rather than outcomes of rigorous empiric research [11••]. There is no work of which these authors are aware that provides empiric guidance for how to address the non-individual determinants of obesity in a clinical setting.

As such, current clinical guidelines for managing obesity focus on the individual-level therapeutic options such as lifestyle modification, medication, and surgery, without mentioning if or how physicians should attempt to address the non-individual (e.g., social, environmental, and macro-level) obesity determinants [12,13,14,15,16]. For example, the National Heart, Lung, and Blood Institute (NHLBI) suggests that the three major components of weight management are diet, exercise, and behavioral therapy [12]. The Endocrine Society, The American College of Physicians (ACP), American Heart Association (AHA), and National Institute for Health and Care Excellence (NICE) all likewise focus on diet and exercise without mentioning how to address the non-individual determinants of obesity [13,14,15,16]. Although these guidelines encourage discussion with patients about barriers to adopting different management strategies, no guideline explicitly discusses methods of addressing obesity’s social, environmental, or macro-level context.

The absence of empiric research or clinical guidelines for how to address the non-individual determinants of obesity is important for a number of reasons. First, with 922.6 million physician’s visits each year, doctors are an important part of the front-line for obesity diagnosis and management [17]. Consequently, failing to equip physicians with information on how to address a patient’s social and environmental context may result in incomplete or unsuccessful obesity management [18•]. Second, given that the social and environmental determinants of heath are thought to contribute to a majority of disease disparities, addressing the non-individual obesity determinants may represent a necessary step to improving obesity health equity in the USA [19]. Finally, physicians and healthcare systems are increasingly held financially accountable for patient outcomes [20•, 21•]. Therefore, understanding and addressing the fundamental causes of disease (i.e., the social and environmental determinants of health) may have important financial implications for clinicians and healthcare systems [22].

Although there is little evidence to suggest how physicians should conceptualize and manage the non-individual determinants of obesity, there is some literature investigating how physicians presently conceptualize and manage obesity. In this paper, we evaluate that literature to better understand physicians’ perspectives on obesity’s social and environmental determinants. We review studies from the past 30 years in order to gather sufficient data to understand how physicians conceptualize and address the social and environmental determinants of obesity in their clinical practices.

Methods

Search Strategy

In 2016, our team of two health service researchers (AN, MP), a medical student (FP), and an academic librarian (DW) searched PubMed, Ovid MEDLINE, and PsycINFO databases for articles that described physicians’ perceptions regarding the causes and management strategies of obesity. Obesity causes were defined as any factor that physicians felt may contribute or lead to obesity. Obesity management strategies were defined as any technique that physicians employed, or thought could be employed, to manage obesity. Our search included all English-language studies published over the past 30 years that addressed variations on the following terms: obesity, attitudes, perceptions, beliefs, knowledge, physicians, adults, causes, etiologies, treatments, and recommendations. Relevant bibliographies were reviewed to ensure comprehensive evaluation of the literature.

Articles that were published over 30 years ago, articles that evaluated the perceptions of non-physicians rather than physicians, and qualitative work or commentaries that were not broadly generalizable representations of physicians’ perspectives were excluded.

Results

Study Characteristics

All studies included in this review were surveys that elicited physicians’ perspectives on the causes and/or management of obesity. Ten of the 22 studies (45%) assessed physicians’ perspective on obesity causes (Table 2); 17 of the 22 studies (77%) assessed physicians’ perspective on obesity management strategies (Table 3). Information on study location and year, respondent characteristics, and overall response rate is presented in Tables 2 and 3.

Table 2 Studies assessing physician perception on the causes of obesity
Full size table
Table 3 Studies assessing physician perception on the management strategies of obesity
Full size table

Almost all studies of obesity causes (90%) queried physicians on each of the individual causative domains of obesity (i.e., biological, psychological, and behavioral). Eight studies (80%) of obesity cause asked physicians about at least one of the non-individual causative domains (i.e., social, environmental), with social domains included in six studies and environmental domains included in five studies. The most common biological causes of obesity discussed in these articles were genetics, hormones, and endocrine/metabolic disorders; the most common psychological causes were stress, depression, anxiety, lack of will power, and low self-esteem; the most common behavioral causes were lack of knowledge about physical activity and proper diet, as well as physical intake of too much food. The most common social factors included in these articles were low income, unemployment, culture, and family influence; and the most common environmental factors were lack of access to healthy food and availability of fattening food in society. No study asked about the macro-level causes of obesity, and no single study evaluated physicians’ perspectives across all possible causative domains of obesity (Table 2).

All 17 studies of obesity management asked physicians about at least one of the individual causative domains (i.e., biological, psychological, or behavioral). The behavioral domain was the most commonly evaluated management strategy domain (n = 17, 100%), followed by biological (n = 13, 76%), and psychological (n = 8, 47%) domains. The most common biological weight loss interventions discussed in these articles were weight loss medication and surgery. The most common psychological weight loss interventions were counseling, hypnosis, and referral to a mental health professional, and the most common behavioral weight loss interventions were advice to eat less, exercise more, or change diet, and referral to a behavioral therapist. Twelve studies (71%) queried physicians on at least one non-individual obesity management domain, most commonly the social domain (n = 12, 71%), followed by environmental (n = 1, 6%) and macro-level domains (n = 1, 6%). The most common social interventions discussed were group weight loss modalities (e.g., support groups, community slimming groups, Weight Watchers), and assessing the role of significant others or family members in obesity. The environmental intervention discussed was assessing the home environment for supportive structure, and the macro-level intervention discussed was policy change. As with studies of obesity causes, no single study of obesity management evaluated physicians’ perspectives across all domains that may affect an individual’s weight status (Table 3).

Several tools were used to elicit physicians’ perspectives on obesity causes and management. Likert scales were used to rate the perceived importance of individual obesity causes (n = 5 studies) and physicians’ level of agreement with statements about obesity causes (n = 4 studies). For example, physicians were asked, “How important is each of the following causes of obesity for your patients?” (answer choices: not at all important (1) to very important (6)) [23,24,25,26,27] or “How strongly do you agree or disagree with the following statements: “Obesity is primarily caused by genetic factors. Obesity is primarily caused by behavioral factors. Obesity is primarily caused by environmental factors.” (answer choices: strongly disagree (1) to strongly agree (6)) [28,29,30,31]. Results from these studies were reported as average Likert scale responses or percentage of physicians in agreement with each statement. In one study, physicians were asked what percentage of their patients was overweight or obese as a result of various potential causes (i.e., “What percentage of patients in your practice is overweight due to: endocrine imbalance; heredity; lack of knowledge; lack of willpower, stress?”) [32]. Results were reported as a percentage of patients.

For obesity management, eight studies asked physicians whether or not the participant physicians recommended various strategies [31, 33,34,35,36,37,38,39]; eight asked how frequently/strongly they recommended various strategies [23, 26, 34, 40,41,42,43,44], and five asked physicians to state the perceived importance/effectiveness of various strategies in managing obesity [24, 30, 35, 38, 42]. Results were reported as average Likert scale responses or as percentage of physician agreement with each statement.

Physician Perceptions about Obesity Causes

When physicians were asked to evaluate the causes of obesity across multiple domains, they rated individual-level causes (i.e., biology, psychology, and behavior) as more important than social or environmental causes in all studies [23, 24, 26, 27]. Similarly, more physicians agreed that individual-level causes were important contributors to obesity than social or environmental causes, in all except one study [28,29,30,31]. This relationship was noticed across multiple physician practice types, multiple countries, and across time.

In many studies, the social and environmental determinants were considered the least important causes of obesity. For example, a 2005 study by Bocquier et al. noted that French general practitioners’ ranked behavioral risk factors (i.e., eats too much fat, 5.2; eats too much, 5.1; eats too much sugar, 4.9; insufficient physical activity, 4.7; and repeat dieting, 4.2) as the most important risk factors for obesity using a 6-point Likert scale, followed closely by individual biological factors (i.e., genetics, 4.5; hormone problems, 3.7), and individual psychological factors (i.e., stress, anxiety, and depression, 4.1) [26]. Social factors in this study (i.e., low income and unemployment, 3.3) were considered the least important risk factors for obesity. Similarly, in a 2013 study conducted in Hungary, Rurik et al. noted that the greatest number of physicians believed that individual behavioral risk factors (i.e., insufficient physical activity, 97%, and eating too much fat, 89%) contributed to obesity, and the lowest number of physicians believed that social risk factors (i.e., low income, unemployment, 46%) contributed to obesity [31].

Even when non-individual causes of obesity (e.g., social, environmental) were not ranked as least important, they were consistently ranked less important than individual (e.g., behavioral, psychological) causes. For example, a 1999 study of American osteopathic physicians found that many physicians perceived behavioral factors (i.e., poor diet, 68%; physical inactivity, 56%) and psychological factors (lack of will power, 39%) to be significant contributors to obesity; fewer physicians ranked social causes (family influence, 11%) or environmental causes (food availability, 18%) as important [23]. Comparable results were seen in more recent studies of American primary care providers and internal medicine resident physicians, as well as British general practitioners [24, 27, 28, 30].

Non-individual causal determinants of obesity only outranked individual causal determinants in one study, in which 92% of primary care providers working at a Los Angeles Veterans Affairs (VA) Hospital reported that people gain weight because “fattening food is too available in our society” (environmental). Fewer physicians in this study reported that biological factors (i.e., “medical conditions” (83%)), behavioral factors (i.e., “fitting physical activity into one’s day” (75%)), or psychological factors (i.e., “lack of willpower” (35%)) contributed to obesity [29].

Physician Perceptions about Obesity Management

When physicians were asked to evaluate potential obesity management strategies across multiple domains, more physicians stated that they used individual-level strategies than social, environmental, or macro-level strategies [33, 35, 36, 38, 43]. Physicians also used individual-level strategies more frequently than they used social, environmental, or macro-level strategies [23, 26, 34, 40, 41, 44] with the exception of one study [30]. Finally, when physicians were asked which management strategies they thought were most effective in the treatment of obesity, they stated that individual-level management strategies were more effective than social, environmental, or macro-level management strategies [35, 38].

These perceptions were also consistent across multiple physician practice types, multiple countries, and over time. For example, a 1991 study by Cade et al. found that British general practitioners were more likely to “always” or “often” recommend individual behavioral changes (i.e., eating less, 78%; exercise, 77%) than social strategies (i.e., Weight Watchers, 54%; family involvement, 32%), for managing obesity [34]. A 2002 study by Fogleman et al. found that Israeli family physicians recommended behavioral management strategies (i.e., increased physical activity, 95%; behavioral treatment, 38%), more often than social strategies (i.e., group support meetings, 25%) [40]. Finally, a 2005 study of French general practitioners found that more than 90% of physicians “often” or “always” addressed individual risk factors, including physical activity, dietary habits, expectations, and psychological state when managing obesity, but only 70–90% “often” or “always” assessed social status [26]. Similar results were found among Australian general practitioners and American osteopathic, military, obstetric, family medicine, and internal medicine physicians.

There was one study in which physicians were more likely to recommend non-individual management strategies than individual strategies. Specifically, when Ogden et al. asked English general practitioners to rank various management strategies of obesity on a five-point Likert scale (from “1 = not at all” to “5 = totally” effective), more physicians (92%) felt social support groups such as Weight Watchers were effective in the management of obesity than biological strategies (i.e., surgery, 40%; medications, 38%) or behavioral/psychological strategies (i.e., counseling, 38%) [30].

Discussion

In this review, we found 22 studies that evaluated physicians’ perspectives on the causes and management strategies of obesity over the past 30 years. Of these 22 studies, 10 evaluated obesity causes and 17 evaluated obesity management. While eight studies of obesity causes (80%) assessed at least one non-individual (e.g., social, environmental) domain of obesity, no single article evaluated physicians’ perspectives across all domains of obesity that might influence weight. Similarly, while 12 studies of obesity management (71%) assessed at least one non-individual obesity management strategy (e.g., social, environmental and macro-level factors), no single study evaluated physicians’ perspectives across all domains that might be used to treat obesity. Additionally, the variety of non-individual obesity causes and management strategies included in these articles was far from comprehensive. Given these limitations in the available literature, it is difficult to know with certainty how physicians comprehensively conceptualize or manage obesity in their clinical practice, and additional research is certainly needed.

However, within the confines of available literature, our review suggests that clinicians generally rank individual characteristics, and in particular behavioral characteristics, as more important in the development and management of obesity than non-individual characteristics. This relationship was seen in 7 of the 8 studies (88%) that evaluated individual and non-individual causes of obesity, and in the 11 of the 12 articles (92%) that evaluated individual and non-individual strategies of obesity management.

Physicians across place (US, France, UK, Hungary, Australia, Israel) and time (30-year time span) ranked individual obesity determinants as more important than non-individual determinants, with only two exceptions: the majority (92%) of internal medicine physicians at a Los Angeles VA Hospital agreed that people primarily gained weight because of environmental factors (i.e., fattening food is too available in our society), and more general practitioners in England (92%) felt that a social weight management strategy (i.e., Weight Watchers) was the solution to obesity than individual management strategies. Of note, this particular social strategy is one with little empiric support.

There are several potential explanations for physician beliefs regarding the causes and management of obesity. First, physicians may simply under-value the impact of social, environmental, and macro-level contexts on obesity, relative to behavioral contexts. McGinnis et al. have described scientific estimates that attribute approximately 40% of an individual's health status to their behavior, 30% to genetics, 20% to social and environmental context, and 10% to healthcare services [45•]. However, this perspective fails to take into account the indirect impact of social, environmental, and macro-level contexts on behavior itself, which magnifies the overall impact of non-individual factors on health [4••]. For example, food choice behavior may be influenced by social (e.g., friends, family, income), environmental (e.g., the presence of grocery stores and fast food restaurants), and macro-level (e.g., media and advertising, the Supplemental Nutrition Assistance Program [SNAP] and other governmental policies) factors, and this overlap is not accounted for in a model which separates behavior from social and environmental influence.

A second potential explanation for physician beliefs about obesity determinants and management may be a lack of perceived ability or responsibility for physicians to address non-individual domains of health within their clinical practice. As discussed, there is a robust literature on clinical interventions that target individual behaviors (e.g., diet changes, physical activity); however, there are relatively few articles and no guidelines that discuss how physicians can or should address the non-individual causes of obesity [11••, 22, 46••]. As such, physicians may consider individual-level influences on obesity to be more amenable to intervention than non-individual influences, or they may feel unprepared to address non-individual determinants in clinical settings.

Traditionally, there has been a division of labor between health care delivery and public health systems. Health care has historically prioritized the needs of individual patients while public health has addressed the needs of populations, including social, environmental, and policy contexts of that population [20•, 47]. However, these distinctions between traditional medical and public health practice are beginning to blur. With recent health policy changes promoting value-based care and population health management [48], health systems are increasingly held accountable for health outcomes and health equity within patient populations [49]. Thus, physicians and health care systems are becoming incentivized to understand and address the non-individual determinants of diseases [50••, 51].

Still, the role of the physician in these health policy changes has been poorly defined, [52••, 53] and a lack of physician understanding for how to comprehensively address the individual and non-individual causes of obesity represents a missed opportunity in the management of obesity and obesity-related diseases (e.g., diabetes, cardiovascular disease). Additionally, given the importance of social and environmental determinants of health (e.g., poverty, limited educational opportunities) in creating health disparities, failing to address these factors has the potential to disproportionately worsen the health of marginalized populations and subsequently increase health disparities [19].

One limitation of this review is that we were unable to identify how much physician respondents considered the influence of social, environmental, and macro-level context on their patient’s behavior when responding to surveys. Thus, there is the possibility that some physicians considered social, environmental, and macro-level influences on their patients’ behaviors, potentially obscuring our understanding of these two answer choices. However, as the articles in our review asked about the effect of social, environmental, and macro-level context separate from diet and exercise, we can only assume that the relative rankings between these two answer choices are independent of one another. Strengths of this review include the breadth of literature on this topic which allowed us to investigate physicians’ perspectives on obesity causes and management across both place and time, as well as the novel look at a previously published literature to better understand a timely question.

Conclusions

Over the past 30 years, no study has investigated physicians’ perspectives on the causes or management of obesity across all possible domains that might influence weight. Nonetheless, our review of existing literature suggests that physicians generally prioritize the importance of individual factors in the development and management of obesity as compared to social and environmental factors. As traditional health care and public health practices move into an era of value-based care and population health management, more research is needed to understand the most appropriate ways for physicians to address the non-individual determinants of obesity.